David H. Moen

Structural changes in multivariate regression models

This study generalizes the work of chin choy and Broemeling (1980) who investigated the change in the regression parameters of univariate linear models. The marginal posterior distributions of the change point, the regression matrices,and the precision matrix are found with the use of a proper multivariate normal-Wishart distribution for the parameters of the model. A numerical study is undertaken in order to gain some insight into the effect that changes in sample size and certain parameter values have on these marginal posterior distributions.

Introducing Expert Systems into the Business School Curriculum

Abstract Many routine problems exist for which a solution has been well defined. When these problems are converted to an expert system, their solutions are more accessible to users. Because of the usefulness of expert systems, they are becoming more popular in the work environment. As a result, all business students should be familiar with their capabilities. This article discusses how teachers can expose students to expert systems by incorporating material on expert system concepts into introductory courses in accounting and statistics at the undergraduate level and in quantitative analysis at the graduate level.

The uncertainty of forecasting: models with structural change versus those without changing parameters

The Bayesian predictive density is found for future observations of the unknown dependent variables for a multivariate linear model with a single shift in the regression matrix. A numerical example shows that it is dangerous to predict future observations with an unchanging parameter model when the appropriate model should include structural change.

Switching linear models: A general approach

ABSTRACT With reference to switching linear models, using the notion of a switching interval, the posterior distributions of all the parameters in the model are obtained. This includes the beginning and end points of the switching interval and the parameters determining the nature of the switch. This is done by studying three cases of the problem: a permanent switch in a finite interval, a permanent switch in an infinite interval, and a temporary switch in a finite interval. The analysis is general in the sense that it can be applied to any problem that can be formulated as a linear model. A numerical study illustrates the methodology.

Bayesian inferences for several autoregressive processes

Several first-order autoregressive processes are studied using Bayesian analysis principles. These autoregressive processes are the same, except perhaps for the level of the process. A test is developed to detect possible level differences by using the mean value functions of these first-order autoregressive processes. To illustrate the test results, a numerical study is conducted using two time series groups.